Cambrian-S: Towards Spatial Supersensing in Video

Native Multimodal Models are World Learners

NANO3D: A Training-Free Approach for Efficient 3D Editing Without Masks

SAPIEN Manipulation Skill Framework, an open source GPU parallelized robotics simulator and benchmark, led by Hillbot, Inc.

Official Implementation of DA²: Depth Anything in Any Direction

Hunyuan 3D Part Segmentation and Generation Pipeline

Hunyuan3D-Omni: A Unified Framework for Controllable Generation of 3D Assets

Voyager is an interactive RGBD video generation model conditioned on camera input, and supports real-time 3D reconstruction.

From Images to High-Fidelity 3D Assets with Production-Ready PBR Material

Efficient Part-level 3D Object Generation via Dual Volume Packing

DeepVerse: 4D Autoregressive Video Generation as a World Model

Mesh-RFT: Enhancing Mesh Generation via Fine-grained Reinforcement Fine-Tuning

[ICML 2025] FreeMesh: Boosting Mesh Generation with Coordinates Merging

Roblox Foundation Model for 3D Intelligence

[ICCV 2025] Official code of DeepMesh: Auto-Regressive Artist-mesh Creation with Reinforcement Learning

High-Resolution 3D Assets Generation with Large Scale Hunyuan3D Diffusion Models.

Official repo for paper "Structured 3D Latents for Scalable and Versatile 3D Generation" (CVPR'25 Spotlight).

Official inference repo for FLUX.1 models

Code for BLT research paper

[NeurIPS 2023] Michelangelo: Conditional 3D Shape Generation based on Shape-Image-Text Aligned Latent Representation

[ICLR 2025] EdgeRunner: Auto-regressive Auto-encoder for Efficient Mesh Generation

[TMLR 2025🔥] A survey for the autoregressive models in vision.

Official code for paper: Scaling Mesh Generation via Compressive Tokenization [CVPR'25]

Tencent Hunyuan3D-1.0: A Unified Framework for Text-to-3D and Image-to-3D Generation

MinHash, LSH, LSH Forest, Weighted MinHash, HyperLogLog, HyperLogLog++, LSH Ensemble and HNSW

[ICCV 2025] From anything to mesh like human artists. Official impl. of "MeshAnything V2: Artist-Created Mesh Generation With Adjacent Mesh Tokenization"

PyTorch implementation of MAR+DiffLoss https://arxiv.org/abs/2406.11838

[NeurIPS'24] NeuRodin: A Two-stage Framework for High-Fidelity Neural Surface Reconstruction

tiktoken is a fast BPE tokeniser for use with OpenAI's models.